Research On Self-Localization In Wireless Sensor Networks

Wireless sensor networks(WSN), as a brand-new technique in information acquisition and processing, can be wildly used in many areas such as Intelligent Transportation Systems(ITS), national defense, environmental supervision, medical treatment, out space exploration and so on. It is considered as one of the most important techniques of the 21st century. Self-Localization is one of the indispensable supporting techniques in wireless sensor networks, and has widely drawn attention of domestic and foreign scholars.In this paper has a deep research on Self-Localization of static WSN. The essential problem of localization error is addressed. The cost of anchor node is hundreds times more than ordinary one, so localization algorithms can't avoid the sparse anchor nodes problem, and the error-propagation problem. For the above two problems, properties of multi-hop-path are analysed and a novel range-free localization algorithm, called DSPL (Difference of Shortest-Path Localization), using the difference of node-to-node hop-counts to determine the distance of unknown node to anchor nodes, is proposed. Thanks to the difference of two short paths, the error-propagation from the unknown node and anchor node is eliminated.DSPL algorithm includes two steps: Step 1: Flooding. Every anchor node send its location to every other node, using dynamic programming, and the least path hop-counts is stored.Step 2: Iteration. At the beginning, every anchor node, send information about their hops to other anchor nodes to neighbors. The unknown nodes use the difference of hop-counts of located neighbors, to calculate the distance to anchors. Using multi-lateration method to get the coordinates. When done, these nodes send their locations to neighbors. Until all the unknown nodes located, the iteration step is done.DSPL2 is proposed, taking the located two-hop nodes into consideration, in order to control the error-propagation using redundant information.Simulation shows, error get more as the hmin value (which means the minimum of all least path hop-counts to anchor nodes) get bigger, the performance of DSPL is superior to the conventional DV-Hop scheme with low error-propagation. DSPL2 reduce error of the nodes with big hmin value.